Project Summary/Abstract. Synthetic RNAs discovered via in vitro selection experiments have wide-ranging applications in biomedical sciences and biotechnology, including therapeutic aptamers that inhibit protein function and ribozymes that control gene expression. The RNA in vitro selection process, however, lacks systematic computational analysis to potentially increase the probability of discovering complex synthetic RNAs. To fill this significant gap, we have developed computational approaches for designing structured RNA pools and optimizing RNA functions to improve the productivity of in vitro selection and directed evolution experiments. In vitro selection experiments and our computational analysis suggest that designed RNA pools possessing diverse structural motifs can enhance discovery of complex RNA motifs which are rarely found in random pools. This is the main hypothesis we aim to demonstrate and apply in this proposal. To address current limitations and develop computational in vitro selection, we aim to develop: (Aim A) computational approaches to structured RNA pool design;(B) methods for screening and testing designed RNA pools;and (Aim C) a computational approach for simulating directed evolution for optimizing RNA functions. In Aim A, we will develop structured pool design approaches that allow generation of user-defined target structures with constant binding or catalytic motifs using a Monte Carlo simulation method. In Aim B, we will computationally test the performance of designed pools using motif scanning and screening (e.g., PI's SVD/TNPACK software tools) methods. In addition, we will experimentally verify that designed pools are superior to random pools via a collaboration with Luc Jaeger, an expert on RNA in vitro selection and nanotechnology. In Aim C, we will develop a computational approach to in vitro evolution by combining the motif scanning/screening methods, the nucleotide transition ("mixing") matrix approach, and the partial least squares method for accumulating beneficial mutations to model RNA motif selection and mutagenic PCR procedures;optimized RNA candidates from our computational in vitro evolution scheme will be experimentally tested by Jaeger's lab. With these algorithmic developments and experimental collaboration, we expect that our computational approaches to pool design and analysis and directed evolution will provide a comprehensive resource to assist experimentalists in designing better in vitro selection experiments and optimizing RNA functions for demanding biomedical applications such as discovering high-binding affinity aptamers targeting proteins in cancer and other diseases. Our project also provides continued excellent interdisciplinary training of students in computational biology, chemistry mathematics, and biomedicine. PUBLIC HEALTH RELEVANCE: Project Narrative Synthetic RNAs discovered via in vitro selection and directed evolution experiments have wide-ranging biological and biomedical applications, including therapeutic RNAs that modulate disease-related proteins. Our work is based on the hypothesis that designed structured RNA libraries are better than random pools for discovering complex RNAs. Our project will develop and test computational approaches for designing structured RNA pools and enhancing directed evolution to assist experimentalists in discovering complex RNAs. We will interact with experimental biomedical researchers to exploit and extend our methods'capabilities for advancing the development of molecular tools for biomedical applications.